A inserção de gotejadores em uma tubulação modifica as linhas de fluxo, provocando turbulência local, que resulta em perdas de carga adicionais maiores que as perdas contínuas na tubulação. Para avaliar a perda total de carga, ao longo da linha lateral de gotejadores, as perdas contínuas e as localizadas, devidas à presença dos emissores no tubo, devem ser consideradas. Este trabalho apresenta os resultados de um experimento conduzido para avaliar as perdas localizadas de carga em gotejadores coaxiais integrados em tubos de polietileno. A perda de carga para diferentes vazões foi determinada em quatro modelos de tubos gotejadores, com sete repetições. Cada segmento de tubo utilizado continha 11 gotejadores. Para cada vazão, a perda localizada de carga foi calculada pela diferença entre a perda de carga no tubo com emissor e a perda de carga continua no tubo uniforme, estimada pela equação de Darcy-Weisbach. Aproximações matemáticas foram sugeridas para calcular a perda de carga com base no coeficiente de carga cinética (K) e em um valor constante de comprimento equivalente (Le). Para cada modelo de tubo gotejador, o coeficiente K foi praticamente independente do número de Reynolds, para R > 10 000, sugerindo que cada gotejador pode ser caracterizado por um valor médio de K, obtido em função de um índice de obstrução. Em decorrência, propôs-se uma alternativa para calcular a perda localizada de carga, provocada por gotejadores coaxiais integrados a tubos de polietileno, em função de relações geométricas entre a área de escoamento através do gotejador e a da seção do tubo. Os resultados encontrados possibilitam concluir que: (a) maior variabilidade dos valores de K ocorreu para os gotejadores não-autocompensados e a maior perda localizada de carga para os autocompesados; (b) o aumento do grau de obstrução (1 - Ag/At) proporcionou o aumento na perda localizada de carga. Considerando a vazão de 20 x 10-5 m3 s-1, nos tubos gotejadores Amanco, observou-se que o aumento de 1,79 vez no grau de obstrução resultou no aumento de 3 vezes na perda localizada de carga; já nos gotejadores Naandan, o aumento de 2,2 vezes no grau de obstrução proporcionou o aumento de 8,9 vezes na perda de carga; (c) a consideração da perda localizada de carga é um critério relevante no dimensionamento de linhas laterais de gotejamento, especialmente quando se utilizam gotejadores autocompensados. A desconsideração da perda localizada de carga levou à superestimativa do comprimento máximo da linha lateral de até 25,7%, para os gotejadores autocompensados, e de 9,5%, para os não-autocompensados; (d) o cálculo da perda localizada de carga, utilizando o modelo potencial para estimar o valor de K em função do índice de obstrução, mostrou bons resultados, comparáveis àqueles obtidos com o valor de K ajustado pelos dados observados em laboratório, o que resultou em pequenas variações na estimativa do comprimento máximo da linha lateral por esses dois procedimentos. Considera-se, assim, que a perda localizada de carga em gotejadores coaxiais, pode ser calculada utilizando-se um coeficiente de carga cinética calculada em função do índice de obstrução. / The insert of the drip in a pipe modifies the flow line, causing local turbulance, which results in more additional head loss than the continued loss in pipes. In order to evaluate the local head loss, along the drip irrigation laterals, the continued and localized loss due the presence of these emitters inside the pipes must be considered. This work shows the results of an experiment carried out to evaluate the local head loss in coaxial drippers integrated to polyethylene pipes. The head loss for different discharges was determined using four drip pipe models, using seven repetitions. Each pipe segment contained 11 drips. For each discharge, the local head loss was calculated by the difference between head loss in the pipe with emitters and the continued head loss in the uniform pipe, estimated by Darcy-Weisbach equation. Mathematical approaches were proposed to calculate the head loss based on the kinetic head coefficient (k) and in a equivalent length of pipe constant value (le). For each drip pipe model, the k coefficient was considered independent from Reynolds number, R >10.000. It can be concluded that each drip can be characterized by an average k, obtained in function from a obstruction index. As a result, it was proposed an alternative to calculate the local head loss caused by coaxial drip integrated to polyethylene pipes, according to geometric relationship between the flow area through the drip and the pipe section. According to the results can be concluded that: (a) the not self-compensated pressure emitters showed more variability in k values and the self-compensated showed greater local head loss; (b) the obstruction index increase (1 ag/at) caused an increase in the total head loss. Considering a discharge 20 x 10 -5 m3 s-1, it was observed in Amanco drip that a 1,79 times increase in obstruction index, resulting in 3 times the head loss and in Naandan drip, it was observed that 2,2 times increase in obstruction index, resulting in 8,9 times the head loss; (c) the head loss is a relevant criterion in the drip laterals design, especially for self-compensated drips. The disregard of the local head loss resulted in a over-estimate of 25,7% in the maximum length lateral, for self-compensated drips, and until 9,5% for not self-compensated drips; (d) the calculation of the local head loss using the potential model to estimate k value, according to the obstruction index, showed good results when comparing to those obtained using k factor adjusted to the observed data in the laboratory. Resulting in few variations in the maximum length lateral estimation for both procedures. It is considered that the local head loss, caused by the integrated drip on drip lateral, can be calculated using kinetic head coefficient according to obstruction index.

Nos projetos hidráulicos de irrigação são contabilizadas as perdas de carga totais, que seriam as perdas contínuas ou principais e as localizadas, objetivando maximizar a uniformidade de distribuição de água, caracterizando um conjunto motobomba adequado ao sistema de irrigação e com isso, minimizando os custos anuais e de implantação do projeto. Com o uso da informática, problemas de cálculos complexos são resolvidos com muita facilidade; desta forma pode-se aplicar modelos mais complexo para calculo da perda de carga nos sistemas de irrigação, resultando em valores mais próximos da realidade, com maiores riquezas de detalhes. A perda de carga representa a dissipação de energia da água em forma de calor, ao longo da tubulação, decorrente da resistência ao escoamento oferecida pela viscosidade do fluido e pela inércia das partículas. É variável de acordo com o tamanho das rugosidades da parede do tubo, diâmetro da tubulação e com a velocidade da água. A indústria de plásticos e seus derivados, com o auxilio da engenharia, tem aprimorado a qualidade dos materiais destinados à fabricação dos tubos, principalmente de polietileno. A utilização de tubos fabricados de matérias plásticos, de menor custo, destinados à irrigação tem aumentado nos últimos anos. A flexibilidade desses tubos traz como consequência o aumento do diâmetro interno com o aumento da pressão, fato este já observado em pesquisa e que não são levados em consideração pelos equacionamentos matemáticos utilizados para determinação da perda de carga. O presente trabalho propõe um modelo onde leva em consideração o módulo de elasticidade (E) do tubo para determinar a alteração do diâmetro em tubos elásticos provocada pela pressão, afetando assim a determinação da perda de carga contínua. Conhecer detalhadamente a causa da perda de energia, com intuito de cada vez mais otimizar a energia gasta por área irrigada no cenário brasileiro, passa a ser de fundamental importância. O Modelo Elástico proposto associado à Equação Universal, apresentou índice de desempenho médio de 0,9 sendo considerado com uma estimativa muito boa da realidade. / Total head losses are accounted in the irrigation hydraulic projects, that would be the continuous losses and the local head losses, aiming to maximize the uniformity of water distribution, characterizing an adequate pump set to the irrigation system e thus, minimizing the project implantation and annual costs. With informatics support, complex calculation problems are solved with ease, therefore it is possible to apply more complex models for head loss calculation in the irrigation system, resulting in values closer to the reality, with greater details. The head loss represents the water energy dissipation as heat, along the piping, due to the resistance to the flow offered by the fluid viscosity and by the particles inertia. It is variable according to the size of the rugosities of the pipe wall, piping diameter and the water velocity. plastic industry and its derivates, with engineering support, have improved the quality of the materials for the pipe manufacturing, mainly polyethylene. The usage of plastic material pipes for irrigation, of lowest cost, has risen in the latest years. The flexibility of these pipes leads to the internal diameter increase with pressure increase, fact already observed in research and that are not taken into account by mathematics equating used to determine the head loss. This paper proposes a model where it takes into account the elastic module (E) of the pipe to determine the diameter alteration in elastic pipes due to the pressure, affecting the determination of continuous head loss. Elastic Module proposed associated to Universal Equation, showed average performance rate of 0,9% being considered a extremely good estimative of reality.

Os estudos focando a otimização do reforço dos tubos de concreto para obras de saneamento são importantes por estes componentes estarem vinculados a grandes necessidades sociais que ainda atingem o Brasil e outros países em desenvolvimento bem como pelo fato destes componentes serem produzidos em série, o que gera economia de escala. Neste sentido, a utilização de fibras como reforço dos tubos trás consigo um grande número de vantagens aplicativas e de desempenho. A tendência internacional de busca por um consenso no dimensionamento e controle do concreto reforçado com fibras (CRF) como material estrutural está sendo direcionada para o fib Model Code 2010. Geralmente, para a caracterização do CRF são utilizados os ensaios de flexão de vigas. Porém, a moldagem desses corpos de prova para o controle do CRF dos tubos apresentou resultados inadequados em estudos anteriores. A definição do reforço ótimo de tubos de CRF através da tentativa e erro no ensaio de compressão diametral do componente deve ser evitada pelos elevados gastos (financeiros, de material e tempo) associados a esta metodologia. O projeto deve ser atingido através de modelos confiáveis que possam otimizar o reforço através da previsão adequada do desempenho do componente. Da mesma forma, deve-se obter procedimentos de controle do material que sejam concatenados com os procedimentos de controle do produto. Neste sentido, esta tese propõe uma metodologia para a caracterização e controle do comportamento mecânico do CRF para a produção de tubos, que permita modelar o comportamento do componente no ensaio de compressão diametral e verificar sua adequação a uma aplicação estrutural. O estudo foi dividido em três fases principais. Na primeira houve a avaliação de um método de ensaio alternativo ao de flexão de prismas para o controle do CRF. Na segunda fase experimental houve a incorporação de modificações no método de ensaio de compressão diametral do componente de modo a aproximá-lo à filosofia do fib Model Code 2010. Nesta etapa foram avaliados tubos com reforço de fibras de aço, macrofibras poliméricas e vergalhões para comprovar a pertinência da nova metodologia de ensaio. Na terceira fase do estudo foi realizada uma simulação numérica para previsão de comportamento dos componentes ensaiados utilizando os parâmetros do material caracterizados através do método de ensaio alternativo validado na primeira fase. Os resultados obtidos na modelagem foram comparados com o resultado experimental do ensaio do componente de modo a validar a metodologia proposta. Os resultados demonstraram a adequação do ensaio Barcelona para a caracterização e controle do comportamento mecânico do CRF destinado à produção de tubos. Essa caracterização poderá subsidiar simulações numéricas do comportamento do componente no ensaio de compressão diametral. Com isto, ábacos de projeto de tubos de concreto reforçados com diferentes tipos de fibras poderão ser desenvolvidos, identificando as classes resistentes que serão atendidas dependendo do diâmetro do tubo e do teor de fibras empregados. Esta tese aponta também para a necessidade de uma revisão na normalização vigente, estabelecendo critérios relacionados ao comportamento pós-fissuração que avaliem o estado limite de serviço e o estado limite último. Isto permitirá uma avaliação homogênea do tipo de reforço e tornará mais adequada a comparação entre distintos tipos de reforço (fibras ou convencional). / The international trend for a consensus about the design and control of fibre reinforced concrete (FRC) as a structural material is based on the fib Model Code 2010. Generally, in order to characterize the FRC, bending tests are used. However, the moulding of these control test specimens of the FRC pipes is not quite simple as has been shown by previous studies. The design of FRC pipes through trial and error in the component-crushing test should be avoided. This test shall be limited to the acceptance control or final validation of a new component, designed through reliable models that optimize the reinforcement and the component performance. These models should allow both design and prediction of the component behaviour related to the quality control. In that sense, this thesis proposes a methodology for characterizing and controlling the mechanical behaviour of FRC for the production of pipes. This study allows modelling the component behaviour in crushing test and verifying their suitability for structural application. The methodology considered consisted in three main topics: evaluation of an alternative method of FRC control test; modification of the procedure of the crushing test method in order to approach the fib Model Code 2010 philosophy, and prediction the mechanical behaviour of the components comparing numerical simulation results with experiments results. In this last topic, the characterization of the materials performance by the alternative test method was considered. The results indicated that the Barcelona test is suitable in order to characterize and control the mechanical behaviour of the FRC used for the production of pipes. This characterization is able to support numerical simulations of the component behaviour in crushing test. It allows the development of design tables identifying the pipes resistant classes considering a variety of parameters such as types and contents of fibres and pipes diameters. This thesis also points out the need for a review of the current standards, establishing parameters related to the post-cracking behaviour to assess the serviceability limit state and the ultimate limit state. This allows a homogeneous evaluation of the reinforcement type and makes it more suited to comparing different types of reinforcement (fibre or conventional).

Este trabalho apresenta um estudo experimental sobre o comportamento de dutos enterrados sofrendo perda de apoio ou elevação em uma determinada região ao longo do comprimento. Foram realizados ensaios com modelos físicos compostos por um maciço de areia pura contendo um tubo repousando sobre um alçapão localizado no centro do vão. A pesquisa envolveu dois programas experimentais distintos. O primeiro foi desenvolvido na Escola de Engenharia de São Carlos/USP e contou com o desenvolvimento e a construção de um equipamento de ensaios possuindo um sistema de alçapão. Os modelos eram dotados de instrumental capaz de medir as deflexões e as deformações específicas ao longo do duto, além das tensões totais no maciço de solo circundante e na base do equipamento. O segundo programa experimental foi conduzido na Universidade do Colorado em Boulder, EUA, e envolveu ensaios em centrífuga. Essa fase da pesquisa teve por finalidade realizar uma investigação visual dos mecanismos de ruptura do sistema composto pelo solo e pelo duto sujeito à perda de apoio. Ambas as etapas do trabalho contaram com a execução de testes com modelos sem tubo. Os ensaios realizados revelaram aspectos importantes do problema investigado. A movimentação ativa ou passiva do alçapão exerceu uma forte influência na redistribuição das tensões no maciço de solo exterior à estrutura, a qual abrangeu distâncias horizontais superiores a 5 B e verticais superiores a 4 B. Após a perda de apoio ou a elevação, o topo, a base e as demais partes do conduto assumiram perfis de deflexão distintos ao longo do comprimento, os quais foram consideravelmente influenciados pela densidade relativa do solo e pela sobrecarga aplicada. A movimentação ativa do alçapão revelou ainda padrões de ruptura incluindo localizações de deformação propagando-se para a região do maciço de solo fora do alçapão / This thesis presents an experimental study on the behavior of buried pipes undergoing a loss of support or elevation in a localized region along its length. Tests with physical models comprising a pure dry sand and a tube resting on a rigid trapdoor base located at the center of its length were performed. The research included two distinct testing programs. The first testing program was carried out at the School of Engineering of Sao Carlos/USP, and included the construction of a laboratory facility containing a trapdoor system. The models were equipped with devices for measuring deflections and strains in the pipe, and total stresses in the soil mass and in the lower boundary of the model. The second phase of this investigation was conducted at the University of Colorado at Boulder, USA, and involved centrifuge testing. The main goal of this part of the research was to assess the failure mechanisms that take place when the pipe experiences loss of support. Models without the pipe were also tested in both phases. Important aspects of the soil-structure interaction were verified with the testing programs carried out in this study. A strong influence on the stress redistribution within the soil mass in the vicinity of the structure was achieved after the active or passive conditions were established, encompassing horizontal distances greater than 5 B and vertical distances greater than 4 B. The crown, the base, and the other parts of the pipe assumed distinct deflection profiles after the loss of support or elevation, which were strongly influenced by the relative density of the surrounding soil and by the surficial applied surcharge. The imposed downward boundary movement included the formation of shear bands initiating at the edge of the void and propagating towards the soil mass in the vicinity of the trapdoor

[pt] Em Influência de defeitos cilíndricos sobre o limite de resistência de anéis de resina éster vinílica reforçada com fibras de vidro é realizado um estudo da influência de defeitos sobre a resistência mecânica de um tubo compósito de resina vinil éster reforçada com fibras de vidro picadas e contínuas. Os defeitos, com geometria cilíndrica, são artificialmente fabricados e têm suas dimensões (raio e profundidade) variados. A resistência mecânica do material foi avaliada através do ensaio de tração de anéis previsto na norma ASTM D 2290. Os resultados são analisados pelo método da ANOVA, obtendo-se, assim, a significância das variáveis relacionadas ao defeito (raio e profundidade). Realiza-se também a modelagem por elementos finitos do ensaio de tração, a fim de se avaliar o comportamento mecânico do material durante o ensaio. / [en] Cilindrical defects influence on the tensile strength of fiberglass reinforced vinyl ester composite rings investigates the influence of defects on the mechanical strength of a fiberglass reinforced vinylester composite pipe. The defects are manufactured in cilindrical shape with dimensions (radius and depth) varying over a selected range. The samples are tested in accordance with the ASTM D 2290 standard and the results are analyzed by the ANOVA method. A finite element model of the test is made in order to investigate the mechanical behaviour of the material during testing.

Offshore on-bottom pipelines are subjected to cycles of thermal and pressure-induced axial expansion, which can cause them to buckle laterally. For an elegant and cost-effective solution, lateral buckling is allowed in a controlled manner. Of the various design parameters, the soil resistance has the greatest associated uncertainty. Previous studies of lateral pipe-soil interaction have used laboratory model tests and continuum-based numerical methods. However, they are economically and computationally expensive, and have mostly been restricted to pipes on undrained clay. To overcome this limitation, this thesis introduces the distinct element method (DEM) as a novel numerical tool for the study of lateral pipe-soil interaction for partially embedded offshore pipelines on sandy seabeds. The DEM directly models the particulate nature of sandy soils, allowing large displacements of discrete bodies and providing insights into the mechanics of the soil at a particle level. Pipe{soil interaction is studied by DEM analyses through four separate research stages: (i) mechanical characterisation of the soil, (ii) specimen preparation and pipeline implementation, (iii) small displacement pipe loading tests and (iv) large displacement pipe loading tests. The soil is modelled as an assembly of spherical particles exchanging contact forces, energy and momentum when they interact. At the microscopic scale, a novel moment-relative rotation contact law is introduced to account for the irregular shape of real sand grains. At a macroscopic scale, the mechanical behaviour of the sand is calibrated using experimental triaxial test data. Additional work includes the numerical preparation of a soil assembly and the implementation of a pipeline object in the open-source DEM code Yade. A novel specimen preparation technique is developed to assemble a homogeneous sample at a desired relative density. The pipeline is implemented as a cylindrical body with a continuously curved surface and a specific mass. Small displacement loading tests are performed, with a segment of the pipeline interacting with a 3D prismatic soil domain, replicating plane strain conditions. The influence of particle size, domain thickness, loading velocity and damping are investigated. The findings provide valuable recommendations for performing DEM simulations of this problem, balancing numerical accuracy and computational effort. Large displacement loading tests are performed to validate the DEM approach and to obtain detailed insights into the nature of the pipe-soil interaction. Monotonic vertical and lateral loading simulations are quantitatively compared with laboratory results. To replicate realistic loading conditions of the pipeline on the seabed, cyclic large displacement tests are also performed. Both the monotonic and the cyclic tests show a good level of agreement with experimental results obtained in previous research. Moreover, the numerical analyses provide insights into the evolution of particle motion and the failure mechanism within the soil.

Cette thèse porte sur l’utilisation de présoutènements pour réduire les tassements, c’est-à-dire la mise en place de boulons au front et d’une voute parapluie au-dessus du front. Elle cherche à fournir des outils pour évaluer l’impact des techniques de construction sur les tassements au moyen de calculs numériques par éléments finis. Une attention particulière est portée aux conséquences des choix de simulation sur les résultats obtenus.Le chapitre 1 présente une comparaison de différentes approches de simulation du creusement d’un tunnel peu profond en déformation plane. Les résultats ne justifient pas de privilégier les approches alternatives par rapport à l’approche classique qui utilise des forces de déconfinement. D’autre part, la disposition des présoutènements impose de réaliser des modélisations tridimensionnelles. Dans toute la suite du mémoire, on se place donc dans le cadre tridimensionnel. Dans un premier temps, le chapitre 2 présente une étude de l’influence de la géométrie du front de taille d’un tunnel sur sa stabilité : on compare le cas d’un front de taille vertical et d’un front de taille incliné ou courbe.Le chapitre 3 porte sur la prise en compte dans les simulations numériques du renforcement du front de taille par boulonnage. Nous avons comparé différentes approches, et montrer l’apport des modélisations qui prennent en compte de manière détaillée l’interaction sol-boulon, pour la simulation d’essais effectués en centrifugeuse.Dans le chapitre 4, on étudie l’influence des boulons au front sur les tassements. Nous avons cherché à discuter l’utilité de décrire en détail certains aspects du processus de construction, comme le cycle de renouvellement des boulons au front. Nous avons également discuté l’influence de la distance de pose du radier en arrière du front.Le chapitre 5 étudie l’influence de la voute parapluie sur les tassements. Nous avons établi une modélisation qui prend en compte la géométrie conique de la voute parapluie et du soutènement et le cycle de renouvellement des tubes de la voute. Enfin, on a proposé une modélisation originale du procédé de réduction des tassements au moyen d’injections réalisées à partir des tubes de la voute, appelé « présoutènement actif ».Les résultats fournissent des éléments pour le dimensionnement des procédés de présoutènements, en montrant, par exemple qu’il existe une valeur limite à partir de laquelle il n’est plus utile d’augmenter le nombre de boulons ou de tubes. Ils montrent également que certains choix de modélisation ont peu d’influence sur les tassements, ce qui justifie de prendre en compte des hypothèses qui simplifient la réalisation pratique des études numériques.La thèse a été financée par le projet FUI Newtun, piloté par Solétanche-Bachy / This thesis focuses on the use of pre-reinforcements to reduce the settlements, namely the reinforcement of the tunnel face by bolts, or the installation of a pipe-roof umbrella placed above the face. It seeks to provide tools to evaluate the influence of the construction techniques on the settlements by means of finite element simulations. A special attention is paid to the consequences of the choices of simulation on the numerical results obtained.Chapter 1 presents a comparison between different approaches of simulation for the excavation of a shallow tunnel in plane strain. The results show that the alternative approaches do not give better results, in terms of width of the predicted settlement trough, than the classical approach based on excavation forces. On the other hand, the implementation of the pre-reinforcements imposes to carry out three-dimensional simulations. In this context, ones discusses, in chapter 2, the influence of the geometry of the tunnel face of a tunnel on his stability: we compare the case of a vertical tunnel face and with an inclined or curved tunnel face.Chapter 3 focuses on the techniques that can be used to take into account in the numerical simulations the reinforcement of the tunnel face by bolting. We have compared different approaches, and showed that models that take into account in more detail the soil-bolt interaction give better results for the simulation of centrifuge tests.In Chapter 4, we discuss the influence of the bolts on the settlements. We have discussed the utility of taking into account in the simulation some aspect of the construction process, such as the renewal of the bolts. Also, we have discussed the influence of the distance between the tunnel face and the invert.Chapter 5 studies the influence of the pipe umbrella on the settlements. We have established a model that takes into account the conical geometry of the pipe umbrella and the lining, and the cycle of renewal of the pipes. In the last place, we have proposed an original modelling of the technique of aiming at reducing the settlements by means of injections carried out from the pipes, called “active pre-reinforcement”.The results provide some elements for the design of the pre-reinforcement elements, and show for example that there exists a threshold value above which it is not useful to increase the number of bolts or of pipes. On the other hand, they show that some choices of simulation have little influence on the computed settlements, which justifies some assumptions that can be useful to reduce the complexity of three dimensional numerical models.The thesis has been funded by the project FUI Newtun (under the coordination of Solétanche-Bachy)

It is recognised through current literature that contaminantin trusion is a primary threat leading to degradation of water quality and threat to human health. The problem is more serious in developing countries where the water supply is intermittent and water distribution system crisscross with sanitary systems. Therefore there is a need to develop the methodology that enables the decision makers and engineers to undertake actions to minimise the risk of contamination of water. The researchs tudy presentedin this thesis addresses these water quality issues by developing the appropriate modelling tools to minimize the risk of contaminant intrusion. The conceptual framework proposed in this study consists of a risk based approach where the process of contaminant intrusion into the systems is traced to know the hazards of contaminant intrusion and vulnerability of the system. The risk of contaminant intrusion into the pipes of a water distribution system is then estimated as the function of hazards and vulnerability. A suit of four models is developed based on this framework. The first model is a water distribution pipe condition assessment model that simulates the potential pathway for contaminant ingress into water pipes by relating it to the deterioration/condition of the pipes. The condition of each pipe is assessed by means of numerous factors related to physical, environmental and operational aspects of the water distribution system. These factors are grouped into different indicators at three levels, depending on the nature of influence of each factor on the deterioration process of the pipe. The uncertainties inherent in these pipe condition indicators are described with fuzzy set theory. A distance based multi-criteria decision making method-fuzzy composite programming has been applied to combine the multilevel pipe condition indicators to form a single indicator to rank the condition of the pipes. The second model is a water flow and contaminant transport modelling tool. This model predicts the envelope of pollution emanating from pollution sources (contaminant zone) and simulates the seepage and contaminant transport in this zone. It is assumed that the seepage of contaminant from pollution sources such as unlined canal/drains and surface water bodies follow saturated flow while from pollution sources such as sewer pipelines, lined canals/drains follow unsaturated flow. Accordingly Richard/Green Ampt equations (unsaturated flow) and Darcy's equation (saturatedf lows) are coupled with advection-diffusion equations that account for water flow and contaminant transport respectively. The third model, the contaminant ingress model, identifies sections of pipe of water distribution system within contaminant zone by combining the outputs from the contaminant seepage model with spatial analysis. The fourth model, the risk assessment model, identifies the risk of contaminant intrusion into a water distribution system from the outputs of the above three models, namely the vulnerability of the water distribution pipe (pipe condition assessment model), the contaminant concentration(contaminant seepage model) and section of pipe in contaminant zone (contaminant ingress model). All these models have been integrated into ArcView GIS to form a decision support system (Improved Risk Assessment of Water Distribution System) and applied to a real water distribution system in Guntur, India for which water pipe network data and data for pollution sources were collected. The modelling results are presented as risk maps that show the potential areas that are under threat of contamination with relative risk scores. It is envisaged that the developed modelling tools will be used by water utilities in developing countries to improve the water quality management by identifying vulnerable areas and understanding threats that exist to the water distribution systems.

Accurate hydraulic conductivity values are necessary for understanding groundwater flow. Methods for estimating hydraulic conductivity show limitations because measured values vary several orders of magnitude in high permeability aquifers. Slug tests, while cost and time efficient, result in values lower than expected. It is proposed that underdamped behavior of water in a well is similar to mass on a damped spring; hence, models constructed to simulate behavior independent of aquifer effects might replicate some tests. The Poiseuille and Darcy-Weisbach models, and extensions of these models considering entry/exit effects, are applied to an aquifer-free laboratory test, and real wells. Aquifer-free laboratory tests are modeled well using both Poiseuille and Darcy-Weisbach models with entry/exit effects. The Poiseuille model for wells does not agree with observed data, possibly because of high Reynolds numbers. The Darcy-Weisbach model does agree with well data significantly better, although the friction factor relies on a single Reynolds number.

Orientadores: Ana Maria Frattini Fileti, Sandra Lúcia da Cruz / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Química / Made available in DSpace on 2018-08-18T11:17:01Z (GMT). No. of bitstreams: 1 Santos_RejaneBarbosa_M.pdf: 7186526 bytes, checksum: a1d9198f8a8453c6c0cc532057d71723 (MD5) Previous issue date: 2011 / Resumo: A cada ano cresce o número de tubulações transportando gases e líquidos, ocasionando a necessidade de aumentar a confiabilidade operacional dos dutos. Considerando a importância do monitoramento dos sistemas de tubulações, o presente trabalho tem por objetivo o desenvolvimento e teste de técnicas de detecção de vazamentos de gases em um sistema de tubulação, baseada nos métodos acústicos, visando à detecção e à determinação da magnitude dos vazamentos, além da localização dos mesmos através da utilização de redes neurais artificiais. Foram provocados vazamentos, com diferentes magnitudes de orifícios, em tubulação de ferro galvanizado de 60m e em tubulação flexível de 100 m de comprimento, transportando ar comprimido. Os ruídos sonoros dos vazamentos foram captados por microfone e analisados para as duas tubulações,sob diversas condições de pressão inicial do ar (2, 4 e 6 kgf/cm2) e de localização do vazamento.O microfone foi instalado no vaso de pressão,localizado no início da tubulação, e acoplado a uma placa de aquisição de dados num microcomputador. Os dados experimentais obtidos através deste microfone foram decompostos em ruídos de diferentes freqüências por meio de um circuito eletrônico de condicionamento de sinais. Para cada pressão inicial, a dinâmica destes ruídos no tempo foi utilizada como entrada do modelo neural para determinar a magnitude do vazamento provocado (saída do modelo). Quando detectada a existência do vazamento pelo primeiro modelo, um segundo modelo neural é acionado para a localização do mesmo na respectiva tubulação. O método de Levenberg- Marquardt com Regularização Bayesiana foi utilizado no treinamento dos modelos neurais.Através dos resultados obtidos observou-se que os vazamentos foram adequadamente detectados por método acústicos, para todos os orifícios utilizados. Entretanto, quando os vazamentos eram muito próximos (1,5 m), os sinais captados pelo microfone utilizado foram muito parecidos, dificultando a localização do mesmo pelo segundo modelo. Os resultados indicaram a grande potencialidade dos modelos desenvolvidos. Para a tubulação flexível, nas pressões 4 e 6 kgf/cm², e para todos os casos da tubulação rígida, os modelos neurais apresentaram 100 % de acerto na detecção. Quanto à predição da magnitude de vazamento, considerando que os dados de saída apresentados à rede neural se constituíam de números inteiros da medida em milímetros, efetuou-se o arredondamento da saída da rede neural e, verificou-se que desta forma, são anulados todos os erros ocorridos em todos os testes realizados. Em testes utilizando apenas a tubulação rígida, com ar comprimido a 6 kgf/cm², o modelo neural para determinar a localização do vazamento desempenhou um bom comportamento, caracterizando a predição do local com erro máximo de 0,6 m. Com o uso desta metodologia elimina-se a necessidade de monitoramento constante do operador humano na tela do computador, analisando os gráficos gerados pelos sinais acústicos, pois os modelos podem gerar um alarme informando a ocorrência, o tamanho e a localização do vazamento / Abstract: Every year the number of pipelines transporting gases and liquids increases, requiring improvements in the operating reliability of pipelines. Considering the importance of monitoring piping systems, the aim of the present work is to develop and test a technique to detect gas leaks in pipes, based on acoustic method, in order to determine the occurrence and the magnitude of leaks, besides locating the leakage, by using neural artificial networks. Leakages of distinct magnitudes were triggered in a 60m-galvanized iron pipe and in a 100m-flexible pipe, which transports compressed air. The audible noise, generated by leaks in both pipelines operating under different initial air pressures (2, 4 and 6 kgf/cm2) and for distinct locations, was captured by the microphone and then analyzed. The experimental data, obtained through a microphone installed inside the pressure vessel and connected to a data acquisition system, was decomposed by an electronic circuit into sounds of different frequencies. For each initial pressure employed, the dynamic of these noises in time was used as input to the neural model in order to determine the occurrence and magnitude of the leak (model output). Once detected the occurrence of a leakage, a second neural model is activated to determine its position in the pipeline. The method chosen for training the neural networks was the Levenberg-Marquardt with Bayesian Regularization. From the results, it was observed that the leaks were properly detected by the acoustic method in all situations. Nevertheless, there were similarities among captured signals from very close leakages (1.5m), which affected the location prediction from the model. The results indicated that the developed models were powerful tools for online monitoring rigid and flexible pipes. For the flexible pipe, using pressures 4 and 6 kgf/cm², and for all operating conditions of the iron pipe, the neural models showed 100% accuracy in leak detection. For the leakage magnitude prediction, whereas the output data presented to the neural network was integer numbers of the measurement (mm), the output of the neural network was rounded, vanishing the small decimal errors detected in all tests. In tests using only the iron pipeline, with compressed air at 6 kgf/cm2, the leak location prediction model performed well, pointing out the local with maximum error of 0.6 m. By using this methodology, the human operator will not need to monitor the graphics on computer screen because the neural models can strike an alarm for the leak occurrence, simultaneously displaying its magnitude and location / Mestrado / Sistemas de Processos Quimicos e Informatica / Mestre em Engenharia Química